批处理之家 - Powered by Discuz! Board

/*
REGEX FIND TOOLS, VERSION 1.0
RF.EXE
COPYRIGHT@2016~2018 BY HAPPY
*/
//静态编译（pcre）
#define PCRE_STATIC
#include "pcre.h"
#include <stdio.h>
#include <string.h>
#include <locale.h>
#include <stdbool.h>
#include <windows.h>
//定义行长（字节）
#define BUFF_SIZE 4096
//检测阈值（字节）
#define CHECK_SIZE 16383
//基础函数群
char* UnicodeToANSI(const wchar_t* Str)
{
int L=WideCharToMultiByte(CP_ACP, 0, Str, -1, NULL, 0, NULL, NULL);
char* Out=(char *)calloc(L+1, sizeof(char));
WideCharToMultiByte(CP_ACP, 0, Str, -1, Out, L, NULL, NULL);
return Out;
}
wchar_t* UTF8ToUnicode(const char* Str)
{
int L=MultiByteToWideChar(CP_UTF8, 0, Str,-1, NULL, 0);
wchar_t* Out=(wchar_t *)calloc(L+1, sizeof(wchar_t));
MultiByteToWideChar(CP_UTF8, 0, Str, -1, (LPWSTR)Out, L);
return Out;
}
bool isUTF8(const char* Str)
{
if(!Str){
return false;
}
const unsigned char* bytes=(const unsigned char *)Str;
while(*bytes){
if(
(
bytes[0]<=0x7F ||
bytes[0]==0x09 ||
bytes[0]==0x0A ||
bytes[0]==0x0D ||
(0x20<=bytes[0] && bytes[0]<=0x7E)
)
){
bytes+=1;
continue;
}
if(
(
(0xC2<=bytes[0] && bytes[0]<=0xDF) &&
(0x80<=bytes[1] && bytes[1]<=0xBF)
)
){
bytes+=2;
continue;
}
if(
(
(bytes[0]==0xE0) &&
(0xA0<=bytes[1] && bytes[1]<=0xBF) &&
(0x80<=bytes[2] && bytes[2]<=0xBF)
) ||
(
(
(0xE1<=bytes[0] && bytes[0]<=0xEC)||
bytes[0]==0xEE ||
bytes[0]==0xEF
) &&
(0x80<=bytes[1] && bytes[1]<=0xBF) &&
(0x80<=bytes[2] && bytes[2]<=0xBF)
) ||
(
(bytes[0]==0xED) &&
(0x80<=bytes[1] && bytes[1]<=0x9F) &&
(0x80<=bytes[2] && bytes[2]<=0xBF)
)
){
bytes+=3;
continue;
}
if(
(
(bytes[0]==0xF0) &&
(0x90<=bytes[1] && bytes[1]<=0xBF) &&
(0x80<=bytes[2] && bytes[2]<=0xBF) &&
(0x80<=bytes[3] && bytes[3]<=0xBF)
) ||
(
(0xF1<=bytes[0] && bytes[0]<=0xF3) &&
(0x80<=bytes[1] && bytes[1]<=0xBF) &&
(0x80<=bytes[2] && bytes[2]<=0xBF) &&
(0x80<=bytes[3] && bytes[3]<=0xBF)
) ||
(
(bytes[0]==0xF4) &&
(0x80<=bytes[1] && bytes[1]<=0x8F) &&
(0x80<=bytes[2] && bytes[2]<=0xBF) &&
(0x80<=bytes[3] && bytes[3]<=0xBF)
)
){
bytes+=4;
continue;
}
return false;
}
return true;
}
//BOM检测
int CheckBom(FILE* fp)
{
unsigned char* buf=(unsigned char*)calloc(3,sizeof(unsigned char));
unsigned char* buf2;
fseek(fp, 0, SEEK_SET);
fread(buf, sizeof(unsigned char), 3, fp);
if(buf[0]==0xEF && buf[1]==0xBB && buf[2]==0xBF){return 3;}
else if(buf[0]==0xFF && buf[1]==0xFE){return 5;}
else if(buf[0]==0xFE && buf[1]==0xFF){return 6;}
else{
fseek(fp, 0, SEEK_SET);
buf2=(unsigned char*)calloc(CHECK_SIZE,sizeof(unsigned char));
fread(buf2, sizeof(unsigned char), CHECK_SIZE, fp);
if(isUTF8(buf2)){
free(buf2);
return 2;
}
free(buf2);
}
return 1;
}
//正则查找函数
int RFindLine(FILE* fp, char* src, int FLAG)
{
bool mode=false;
int BOM=0, EN=0, i=0, n=0;
FILE* sp;
char* Li=(char *)malloc(BUFF_SIZE*sizeof(char));
char* LineV;char* LineU;
pcre *re;
int erroffset, ovector[30], rc;
const char *error;
if ( (FLAG&0x0F)==0x02 ){
pcre_compile(src, 0, &error, &erroffset, NULL);
if( (re=pcre_compile(src, 0, &error, &erroffset, NULL)) == NULL ){
fputs("PCRE compilation failed", stderr);
exit(1);
}
}else if( (FLAG&0x0F)==0x03 ){
if( (sp=fopen(src, "rb"))==NULL ){
fputs("Read matching failed", stderr);
exit(1);
}
}
//BOM偏移值
BOM=CheckBom(fp);
if (BOM==1 || BOM==2){
EN=0;
}else if(BOM==5 || BOM==6){
EN=2;
}else if(BOM==3){
EN=3;
}
//执行偏移值
fseek(fp, EN, SEEK_SET);
//执行匹配过程
if(BOM==1){
char* Line=(char *)malloc(BUFF_SIZE*sizeof(char));
while(!feof(fp)){
memset(Line, 0, BUFF_SIZE*sizeof(char));
fgets(Line, BUFF_SIZE, fp);
i++;
if ( (FLAG&0x0F)==0x01 ){
if( strstr(Line, src)!=NULL ){
mode=true;
}else{
mode=false;
}
}else if( (FLAG&0x0F)==0x02 ){
if( pcre_exec(re, NULL, Line, strlen(Line), 0, 0, ovector, 30) >= 0 ){
mode=true;
}else{
mode=false;
}
}else if( (FLAG&0x0F)==0x03 ){
mode=false;
fseek(sp, 0, SEEK_SET);
while(!feof(sp)){
memset(Li, 0, BUFF_SIZE*sizeof(char));
fgets(Li, BUFF_SIZE, sp);
char* tp=Li;
while(*tp=='\t' ||*tp==' ' ||*tp=='\r' ||*tp=='\n'){tp++;}
int tp_LEN=strlen(tp);
tp[tp_LEN-2]=(tp[tp_LEN-2]=='\r')?'\0':tp[tp_LEN-2];
if(tp[0]!='\0' &&strstr(Line, tp)!=NULL){
mode=true;
break;
}
}
}
//输出显示
if( (FLAG>>4)==0x03 && mode==true ){
fprintf(stdout, "%d:%s", i, Line);
}else if(
((FLAG>>4)==0x02 && mode==false)||
((FLAG>>4)==0x01 && mode==true )
){
fputs(Line, stdout);
}
}
}else if(BOM==2 || BOM==3){
char* Line=(char *)malloc(BUFF_SIZE*sizeof(char));
while(!feof(fp)){
memset(Line, 0, BUFF_SIZE*sizeof(char));
fgets(Line, BUFF_SIZE, fp);
i++;
if(BOM>1){LineU=UnicodeToANSI(UTF8ToUnicode(Line));}
if ( (FLAG&0x0F)==0x01 ){
if( strstr(LineU, src)!=NULL ){
mode=true;
}else{
mode=false;
}
}else if( (FLAG&0x0F)==0x02 ){
if( pcre_exec(re, NULL, LineU, strlen(LineU), 0, 0, ovector, 30) >= 0 ){
mode=true;
}else{
mode=false;
}
}else if( (FLAG&0x0F)==0x03 ){
fseek(sp, 0, SEEK_SET);
while(!feof(sp)){
memset(Li, 0, BUFF_SIZE*sizeof(char));
fgets(Li, BUFF_SIZE, sp);
if( strstr(LineU, Li)!=NULL ){
mode=true;
break;
}
}
}
//输出显示
if( (FLAG>>4)==0x03 && mode==true ){
fprintf(stdout, "%d:%s", i, LineU);
}else if(
((FLAG>>4)==0x02 && mode==false)||
((FLAG>>4)==0x01 && mode==true )
){
fputs(LineU, stdout);
}
}
}else if(BOM==5){ //Unicode
wchar_t* LineW=(wchar_t *)calloc(BUFF_SIZE, sizeof(wchar_t));
while(!feof(fp)){
memset(LineW, 0, BUFF_SIZE*sizeof(wchar_t));
fgetws(LineW, BUFF_SIZE, fp);
i++;
LineV=UnicodeToANSI(LineW);
if ( (FLAG&0x0F)==0x01 ){
if( strstr(LineV, src)!=NULL ){
mode=true;
}else{
mode=false;
}
}else if( (FLAG&0x0F)==0x02 ){
if( pcre_exec(re, NULL, LineV, strlen(LineV), 0, 0, ovector, 30) >= 0 ){
mode=true;
}else{
mode=false;
}
}else if( (FLAG&0x0F)==0x03 ){
fseek(sp, 0, SEEK_SET);
while(!feof(sp)){
memset(Li, 0, BUFF_SIZE*sizeof(char));
fgets(Li, BUFF_SIZE, sp);
if( strstr(LineV, Li)!=NULL ){
mode=true;
break;
}
}
}
if ( (FLAG>>4)==0x03 && mode==true ){
fprintf(stdout, "%d:%s", i, LineV);
}else if(
((FLAG>>4)==0x02 && mode==false)||
((FLAG>>4)==0x01 && mode==true )
){
fputs(LineV, stdout);
}
}
}else if(BOM==6){ //Unicode big endian
wchar_t* LineW=(wchar_t *)calloc(BUFF_SIZE, sizeof(wchar_t));
while(!feof(fp)){
memset(LineW, 0, BUFF_SIZE*sizeof(wchar_t));
fgets(LineW, BUFF_SIZE, fp);
i++;
for(n=0;LineW[n]!=0x0000;n++){
LineW[n]=(LineW[n]&0x00FF)<<8|(LineW[n]&0xFF00)>>8;
}
LineV=UnicodeToANSI(LineW);
if ( (FLAG&0x0F)==0x01 ){
if( strstr(LineV, src)!=NULL ){
mode=true;
}else{
mode=false;
}
}else if( (FLAG&0x0F)==0x02 ){
if( pcre_exec(re, NULL, LineV, strlen(LineV), 0, 0, ovector, 30) >= 0 ){
mode=true;
}else{
mode=false;
}
}else if( (FLAG&0x0F)==0x03 ){
fseek(sp, 0, SEEK_SET);
while(!feof(sp)){
memset(Li, 0, BUFF_SIZE*sizeof(char));
fgets(Li, BUFF_SIZE, sp);
if( strstr(LineV, Li)!=NULL ){
mode=true;
break;
}
}
}
if ( (FLAG>>4)==0x03 && mode==true ){
fprintf(stdout, "%d:%s", i, LineV);
}else if(
((FLAG>>4)==0x02 && mode==false)||
((FLAG>>4)==0x01 && mode==true )
){
fputs(LineV, stdout);
}
}
}
fflush(stdout);
if( (FLAG&0x0F)==0x02 ){pcre_free(re);}
free(Li);
return 0;
}
//帮助信息
void Help_Info(FILE* stream, int Exit_Code)
{
fprintf(stream,
"COPYRIGHT@2016~2018 BY HAPPY, VERSION 1.0\n"
"REGEX FIND TOOLS\n"
"-----------------------------------------------------------------------------\n"
"rf [/F|/N|/V] [/S:[match string] ]|\n"
" [/R:[pcre expression] ]|\n"
" [/G:[ANSI strings file]] [txtfile]\n"
"-----------------------------------------------------------------------------\n"
" /H Show help information\n"
" /F Finds the line to which matches\n"
" /N Print the line number\n"
" /V Shows all rows that do not contain matching regulars\n"
" /S: Finds the line to which the string matches\n"
" /R: Finds the line to which the regular expression matches\n"
" /G: Gets the matching strings from a ANSI strings file\n"
"-----------------------------------------------------------------------------\n"
" 11/06/2016\n"
);
exit(Exit_Code);
}
//主函数入口
int main(int argc, char** argv)
{
FILE* fp;
unsigned char FLAG=0;
if( (argc==4) && (argv[1][0]=='/') && (argv[2][0]=='/') && (argv[2][2]== ':') ){
switch(argv[1][1]){
case 'F':
case 'f':
FLAG|=0x10;
break;
case 'V':
case 'v':
FLAG|=0x20;
break;
case 'N':
case 'n':
FLAG|=0x30;
break;
default:
Help_Info(stderr, 2);
}
switch(argv[2][1]){
case 'S':
case 's':
FLAG|=0x01;
break;
case 'R':
case 'r':
FLAG|=0x02;
break;
case 'G':
case 'g':
FLAG|=0x03;
break;
default:
Help_Info(stderr, 1);
}
}else {
Help_Info(stderr, 3);
}
if( (fp=fopen(argv[3], "rb"))==NULL ){
fputs("Read failed", stderr);
return 3;
}
RFindLine(fp, argv[2]+3, FLAG);
fclose(fp);
return 0;
}